Schizophrenia is a highly disabling illness that impacts 0.5-1% of Americans. The disability of the illness is more strongly associated with cognitive deficits and negative symptoms than it is with the positive symptoms of the illness including hallucinations and delusions. Negative symptoms are generally defined as the absence of normal function, but the actual mechanisms involved in generating this absence have remained unknown, thereby stifling rational treatment development. The overarching goal of this application is to fundamentally alter the understanding of negative symptoms by rigorously testing a highly specific hypothesis about the origins of avolition/anhedonia in people with schizophrenia. This hypothesis has been formalized in a computational model that suggests that people with schizophrenia have a deficit in the ability to represent the positive expected value of stimuli and response alternatives, coupled with an intact ability to learn from aversive outcomes. This deficit in representing value is also thought to lead to reduced exploration of behavioral alternatives when uncertain about the likely payoffs of different choices. The project uses a program of behavioral experiments to test this hypothesis in the areas of learning from outcomes and decision making. In addition, we will explore the relationship between this deficit and current cognitive psychological models of the causes of negative symptoms, as well as the importance of this deficit for the prediction of successful outcome from a behavioral treatment approach that uses reinforcement to shape behavior. This computational approach also leads to a highly specific hypothesis about the neural mechanisms that are implicated in a deficit in representing expected value. To address this hypothesis, we will take advantage of the temporal resolution of EEG to test whether abnormalities in neural activity occur at the time of decision, as predicted, or instead occur at the time of feedback delivery, as would be expected if patients were unable to use the dopamine system to signal positive prediction errors when outcomes are better than expected. The goal of both the behavioral and neurophysiological studies is to provide an explicit, mechanistic understanding of negative symptoms and evaluate the application of this approach to current treatment approaches. Because reward circuitry is highly conserved across mammalian species, it should be possible to back translate to the animal models needed to guide drug development research.